Murine retroviruses carry three genes (gag, pol, env) encoding polyproteins essential for virus formation and infection. The gag gene encodes structural proteins for virus inner capsid formation, pol provides the enzymatic proteins needed for processing viral proteins, for reverse transcription of packaged RNA molecule(s) and finally for integration of the provirus. The env gene encodes a polyprotein which is transferred via the endoplasmatic reticulum, proteolytically processed and glycosylated by cellular enzymes and finally integrated into the cellular membrane. It consists of different domains: A signal peptide which is cleaved off after transport to the endoplasmatic reticulum, an extracellular surface domain (SU) and a transmembrane domain (TM) which is both noncovalently and via S bridges bound to SU and thereby anchoring the protein complex within the cellular membrane. SU is cleaved from TM by a cellular protease during transport/processing of the polyprotein through the secretory pathway.
In a final step of virus release from the cell, retrovirus particles are formed by assembly of gag and gagpol proteins at sites of the cellular membrane which are covered with env proteins; it has been shown for certain virus types that the intracellular part of TM is interacting with the MA domains of gag polyproteins (Brody et al., 1994; Dorfman et al., 1994) and is thereby involved in the specific integration of env proteins. Subsequently, budding of virus capsids takes place through the cellular membrane. As a result, retrovirus particles are enveloped by cellular membrane and spiked with homotrimers of env proteins. While retroviral env proteins are highly represented in virus membranes, a similar abundance has not been shown for any cellular surface marker. Thus, in contrast to specific env the incorporation of cellular proteins into virus outer membrane is assumed to be random.
During the process of infection, the SU protein is required for interaction with the respective receptor while the TM protein is responsible for subsequent fusion of the virus particle with the interacting cell.
For gene transfer modified retroviruses are used which have been immobilised and therefore cannot leave the host cell. However, they still express the respective viral genes. In these so-called packaging or helper cells the retroviral transcription unit has been modified by deletion of the packaging sequence required for specific incorporation of viral RNA into the virus particles, a sequence that provides transmission of the retroviral genome in wild type viruses. By transfection of packaging cells with retroviral vectors carrying the respective therapeutic gene and the packaging sequence recombinant virus particles are formed. Since the viral genes themselves have been immobilized, only the vector RNA is packaged into the virus particle and transduced to the infected cell.
During the last years, much effort has been undertaken to modify the host range of retroviruses by alteration of the retroviral env protein. Many attempts have been undertaken to create functional, i.e. infectious fusion proteins for retroviruses. Most of them failed. Several groups could show that N-terminal tagging of SU is possible. Although redirection of viruses to other cellular receptors has not been achieved with murine retroviruses to date, in some cases a tag on env does not inhibit receptor interaction.
One of the mechanisms of human defense invading organisms attack is the complement mediated lysis of invading organisms. The complement mediated inactivation is a multistep enzymatic cascade which finally results in formation of a membrane attack complex (MAC) mediating the perforation of membranes and subsequent lysis of the invading organism. It is either initiated by antigen-antibody complexes (classical pathway) or via an antibody independent pathway which is activated by certain particular polysaccharides, viruses and bacteria (alternative pathway).
Human organs and cells themselves are protected to complement mediated lysis. This protection is achieved by expression of so-called complement inactivation factors. So far, five human factors are known; CD35 (CR1) is released from the cells and acts mainly extrinsically. In contrast, CD59, CD46 (MCP), CD55 (DAF) and HRF are integrated into the cellular membrane: CD46 (MCP) is a classical transmembrane protein while HRF, CD59 and CD55 are GPI-anchored. These factors can interrupt the complement cascade at two different stages: DAF, CR1 and MCP act at an early stage of both the alternative and the classical pathway; in contrast, CD59 and HRF inhibit the assembly of the membrane attack complex which is the final step of both pathways resulting in channel formation and lysis.
Viruses with a natural human host range have developed strategies to overcome this defense system. So for Herpes Saimiri it has been reported that this virus has overcome complement mediated lysis by presenting a virus encoded complement control protein homolog (CCPH) (Fodor et al., 1995), which is also able to block the cascade at the early stage. HIV can incorporate cellular DAF, MCP, and CD59 at levels that protect from complement-mediated destruction (Saifuddin et al., 1995; 1997). 
Early investigations provided evidence that murine retroviruses are inactivated via the complement mediated attack (Welsh et al., 1975). At that time, evidence was given that the cascade is started by direct binding of the extracellular domain of the TM domain to C1, the first enzyme of the classical pathway, resulting in a direct, immunoglobulin independent activation of the classical pathway (Cooper et al., 1976; Bartholemew et al., 1980). Meanwhile, an additional reason for complement instability of mouse retroviruses became apparent (Rother et al., 1995; Takeuchi et al., 1996): natural human antibodies directed against the xcex1-galactosyl epitope presented by non-old world primate cells have been shown to be responsible for complement mediated inactivation via the alternative pathway. Thus, the glycosylation pattern of the host cell is believed to contribute to the complement sensitivity of retroviruses generated thereof. Indeed, murine retroviruses produced from certain human cell lines are stable against complement mediated attack (Cosset et al., 1995). However, in spite of this altered glycosylation, not all human cells are suitable to provide this protection (Takeuchi et al., 1994) indicating that the problem of complement mediated lysis cannot be not exclusively deduced to this specific glycosylation pattern. Thus, TM mediated activation of complement and/or other, still unknown mechanism might contribute to the formation of complement sensitive viruses even from human cells.
Thus, the invention concerns a ssDNA or dsDNA encoding a fusion protein of
a retroviral surface protein (ENV) or a part thereof comprising its transmembrane domain and
a catalytically active domain of a complement inactivation factor.
As regards an ENV part, it should be kept in mind, as already mentioned before, that an extracellular surface domain (SU) is required for interaction with the respective receptor; cf. also Examples on FIGS. 1 to 4.
The ssDNA or dsDNA according to the invention can be characterized in that the catalytically active domain of the complement inactivation factor is fused with its N-terminus to the C-terminus of the retroviral surface protein (ENV) or its part.
Further, the ssDNA or dsDNA according to the invention can be characterized by a human complement inactivation factor, especially DAF, CD 59, MCP or HRF; or a virus complement inactivation factor, especially CCPH.
Further, the ssDNA or dsDNA according to the invention can be characterized by a retroviral SU protein, a retroviral TM protein or parts of these protein, especially amphotropic ENV 10 A 1, ENV 4070 A, ecotropic ENV, GALV ENV or chimeras thereof.
Concerned is further a ssDNA or dsDNA according to the invention which offers a virus or a cell, especially a packaging cell, stability against a complement system, especially a human complement system, and is obtainable by
subjecting viruses or cells, especially packaging cells, which are characterized by a fusion protein as defined by any of the preceding claims to a complement system, especially a human complement system,
selecting a virus or a cell which remains stable and
identifying and isolating said ssDNA or dsDNA from said virus or cell.
Further, the invention concerns a fusion protein encoded by a ssDNA or dsDNA according to the invention.
Further, the invention concerns a virus which is transduction efficient and stable against the human complement system, characterized by at least one fusion protein according to the invention as envelope protein.
The virus according to the invention can be characterized in that the virus is a retrovirus, especially a murine virus, is a virus other than a retrovirus, is a composite virus or is a pseudotyped virus.
Further, the invention concerns a packaging cell for the production of a virus according to the invention, characterized by an integrated or unintegrated dsDNA according to the invention.
Further, packaging cell according to the invention can be characterized in that it is a human cell or a non-human cell, especially a murine cell.
Further, the packaging cell according to the invention can be characterized in that it comprises in addition a dsDNA encoding a wild-type retroviral envelope protein or a part thereof comprising its receptor interacting domain.
Further the packaging cell according to the invention can be characterized in that it is derived from a known packaging cell, especially PA 317 or GP+AM 12.
Further, the invention concerns a use of a virus according to the invention for the enhancement of the transduction efficiency of viruses for human gene therapy.
Further, the invention concerns a use of a fusion protein according to the invention for the generation of complement resistant cells for transplantation.
Further, the invention concerns a use of a fusion protein according to the invention for the reduction of the risk of a formation of replication competent viruses by means of packaging cells via recombination by eliminating the intrinsic sequence overlap of retroviral gagpol and env genes.
Further, the invention concerns a use of a fusion protein according to the invention for generating complement stabilized packaging cells by coexpression of wild-type ENV proteins and fusion proteins according to the invention.
Further, the invention concerns a use of a fusion protein according to the invention for generating complement stabilized viruses according to the invention.